石材表面的线类缺陷检测  

Lines defect detection on the surface of the stone

作　　者：赵民[1] 武晓龙[1] 

机构地区：[1]沈阳建筑大学机械工程学院,沈阳100168

出　　处：《现代制造工程》2017年第2期113-116,共4页Modern Manufacturing Engineering

基　　金：住建部项目(2012-K3-44)

摘　　要：在单一颜色的石材大板中,裂纹是主要的一种线性缺陷。在数字图像处理范畴内对石材大板裂缝的识别算法进行研究。针对裂缝图像的特点,首先对图像进行预处理,以便保证识别的准确性。重点对图像校正以及图像增强进行了研究,达到消除噪声的同时又能最大程度上保护好裂纹边缘的目的。然后采用改进后的阈值分割法对线性裂纹滤波图像进行分割;采用数学形态对裂纹进行单像素细化,获得裂纹尺寸。In a single color stone slab, crack is one of the main linear defects. Conducts the research in digital image processing in the category of stone slab crack recognition algorithm. According to the characteristics of fracture image, first of all is to do the im- age preprocessing,in order to ensure the accuracy of recognition. Mainly study is image correction and enhancement,in order to e- liminate noise and to protect hounding around the largest degree. Then image of linear fracture filtering was segmented by using the improved threshold segmentation algorithm;Mathematical morphology was used to single pixel refinement of crack, and to gain crack size.

关 键 词：线性裂纹 识别算法 裂缝检测 

分 类 号：TH16[机械工程—机械制造及自动化]